Finite Element Model Loading (please post your opinion)

[MarcoE]

I am currently working a CFD analysis for a large section of ductwork.

The project is a team effort, which comprises of teams here in the States and one in the UK.

Given our duct model, it can be loaded from the front or the back. The front has a larger area such that if its loaded from the front a max of 3 m/s is used, and if loaded from the back 20 m/s is applied to an extended exit of the duct. The results for the velocity maximums are similar. The one loaded from the front has a velocity max of 45 m/s and the back loaded model produces a max of 43.5 m/s.

HERE IS THE QUESTION (to settle a dispute between colleages):

should the front loaded or rear loaded model be used and why do you think so?


Here's my humble opinion. In the design of this ductwork both loadings are similar to what we have in real life, but the frontloaded model should be used because it produces the maximum and that is what we design for the "worst case" senario.

Your comments are appreciated.

Marco Escuandolas

 

[rastogisk]

My answer is that the analysis must reproduce the real application.

If the real application uses front load you should analyze as front loaded on the other hand if the real application uses back load then you should analyze it as back loaded.

Even if the two cases produce similar results, it saves time to convince other people that it does not matter as to how you load the model.

I hope it helps.

 

[Mrmech]

hai!
i agree with rastogisk. Computer simulation is man made and it sholud represent real problem to get valid answer.

mrmech

 

[corus]

Looking at the maximum value produced is not always a good guide to which method of loading a model is the best. In a simple mechanical case reaction forces at restraints are equivalent to the input forces and whether you apply the input load at one end of an object and restrain the other, or vice versa, makes no difference mathematically. However, you might find that one method gives a better distribution of results than the other in the region that you are interested in perhaps due to some uncertainty in the application of the load or restraint. I'd look at the results and make a judgement on which of the two looks more realistic at the region you are interested in rather than which model gives you the biggest number, which is not a worst case scenario.

 

[MarcoE]

Thanks to all of you for your thoughts on this.